Sentiment analysis of your documentation As A Service (tm).
This #AstroHackWeek community hack project uses documentation sentiment analysis via the python natural language toolkit (NLTK) package to analyze the astropy docs. Our goal is to use the results to improve astropy documentation and provide a tool for other software package developers to use more user-friendly language in their docs.
Some intermediary goals are:
- Compare sentiments of subdocs of subpackages with in astropy to see if they are uniform (probably not!) and explore the differences
- Use the results to poll Astro Hack Week participants and define what kinds of documentation language are the most user-friendly
- Write a script that can run on any Sphinx documentation build to analyze user-friendliness
- Based on the output of the script, provide concrete feedback to developers about how their documentation could be changed to be more user-friendly
Adrian Price-Whelan @adrn, Meredith Rawls @mrawls, Ellie Schwab @elliesch, Brigitta Sipocz @bsipocz